With the constant increasing of the functions of the intelligent terminal and the portable device, the voice recognition technology is increasingly applied extensively; for example, the voice recognition function is very practical during driving or when the hands need to do other things.
According to the current situation, most mobile terminals with the voice recognition function provide the key or the button to the user to start the voice recognition procedure; the user can send the voice instruction after pressing the key or the button, and the terminal makes the corresponding operation according to the voice instruction of the user. However, there is a very big problem to do it in this way that the user still needs to rely on the hands to click the mobile terminal.
In order to further free the hands of the user, the voice recognition function supporting the voice wake-up technology also appears at present. After the voice wake-up technology is used, the terminal is in the audio detection state all the time, and once special voice wake-up instruction (such as, Hey, voice assistant) is detected, then the voice recognition function is started to identify the voice instruction of the user. The voice wake-up technology can further free the hands of the user, however, because the terminal will consume the electric quantity of the battery to monitor the voice wake-up instruction, the mobile terminal supporting the voice wake-up technology has the problem that the electric quantity is consumed fast.
Therefore, how to make the voice recognition function of the mobile terminal not only free the hands of the user to the greatest extent but also save the power consumption is a technical problem required to be solved.